summaryrefslogtreecommitdiff
path: root/src
AgeCommit message (Collapse)Author
2011-02-16r300g: disable linear filtering for float texturesMarek Olšák
2011-02-15glsl: Reinstate constant-folding for division by zeroChad Versace
Fixes regression: https://bugs.freedesktop.org/show_bug.cgi?id=34160 Commit e7c1f058d18f62aa4871aec623f994d7b68cb8c1 disabled constant-folding when division-by-zero occured. This was a mistake, because the spec does allow division by zero. (From section 5.9 of the GLSL 1.20 spec: Dividing by zero does not cause an exception but does result in an unspecified value.) For floating-point division, the original pre-e7c1f05 behavior is reinstated. For integer division, constant-fold 1/0 to 0.
2011-02-15Revert "glsl: Fix constant-folding for reciprocal expressions"Chad Versace
This reverts commit b3cf92aa916ee0537ee37723c23a9897ac9cd3e0. The reverted commit prevented constant-folding of reciprocal expressions when the reciprocated expression was 0. However, since the spec allows division by zero, constant-folding *is* permissible in this case. From Section 5.9 of the GLSL 1.20 spec: Dividing by zero does not cause an exception but does result in an unspecified value.
2011-02-15tnl: Add support for datatype GL_FIXED in vertex arraysChad Versace
Before populating the vertex buffer attribute pointer (VB->AttribPtr[]), convert vertex data in GL_FIXED format to GL_FLOAT. Fixes bug: http://bugs.freedesktop.org/show_bug.cgi?id=34047 NOTE: This is a candidate for the 7.9 and 7.10 branches.
2011-02-16r600g: add srgb compressed formats to the list.Dave Airlie
2011-02-15gallium: add usage parameter to pipe_buffer_createMarek Olšák
And set a reasonable buffer usage flag everywhere instead of just PIPE_USAGE_DEFAULT.
2011-02-15r600g: Fix RGB10_A2 format handlingFabian Bieler
2011-02-15r600g: fix regression in r6/7xx since mipmap reworkDave Airlie
I typod this when copy-pasting.
2011-02-15nvc0: Fix uninitialized variable warning.Vinson Lee
Fixes this GCC warning. nvc0_tgsi_to_nc.c: In function 'bld_tex': nvc0_tgsi_to_nc.c:1392: warning: 'dim' may be used uninitialized in this function
2011-02-15r300g: offload the CS ioctl to another threadMarek Olšák
This is a multi-threading optimization which hides the kernel overhead behind a thread. It improves performance in CPU-limited apps by 2-15%. Of course you must have at least 2 cores for it to make any difference. It can be disabled with: export RADEON_THREAD=0
2011-02-15r600g: add support for s3tc formats.Dave Airlie
On r600, s3tc formats require a 1D tiled texture format, so we have to do uploads using a blit, via the 64-bit and 128-bit formats Based on the r600c code we use a 64 and 128-bit type to do the blits. Still requires R600_ENABLE_S3TC until the kernel fixes are in, this has only been tested on evergreen where the kernel doesn't yet get in the way.
2011-02-15r600g: fix miptree calculationsDave Airlie
the miptree setup and pitch storing didn't work so well for block based things like compressed textures. The CB takes blocks, where the texture sampler takes pixels, and transfers need bytes, So now we store blocks/bytes and translate to pixels in the sampler. This is necessary for s3tc to work properly.
2011-02-15r600g: drop tiled flagDave Airlie
we can work this out from the array_mode and it makes more sense to do that. Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-15st/mesa: fix compressed mipmap generation.Dave Airlie
If the underlying transfer had a stride wider for hw alignment reasons, the mipmap generation would generate badly strided images. this fixes a few problems I found while testing r600g with s3tc Signed-off-by: Dave Airlie <airlied@redhat.com>
2011-02-15st/mesa: fix GLES buildMarek Olšák
Broken since d5062fb3a315c46d77d5c954a3e3c14be1907d33. I wonder why this code is hidden behind #if by default.
2011-02-15r300g: actually implement the is_buffer_busy hook the right wayMarek Olšák
Ooops.
2011-02-15r300g: handle interaction between UNSYNCHRONIZED and DONTBLOCK flags in bo_mapMarek Olšák
The VBO module uses both, but they are somewhat opposite to each other. In this case, we pick UNSYNCHRONIZED and ignore DONTBLOCK.
2011-02-15r300g: fix a possible race condition when mapping a bufferMarek Olšák
This is the last one I think.
2011-02-15r300g: implement fences using dummy relocationsMarek Olšák
So finally we have them.
2011-02-15r300g: fix SIGFPE on debug buildsMarek Olšák
2011-02-15r300g: inline some of the pipe_buffer_map/unmap callsMarek Olšák
2011-02-14r300g: do not track whether occlusion queries have been flushedMarek Olšák
The winsys takes care of flushing automatically.
2011-02-14r300g: flush CS in bo_map even if we get USAGE_DONTBLOCKMarek Olšák
Because an app may do something like this: while (!(ptr = bo_map(..., DONT_BLOCK))) { /* Do some other work. */ } And it would be looping endlessly if we didn't flush.
2011-02-14st/python: remove pipe_vertex_buffer::max_indexVinson Lee
This is a follow-up to commit cdca3c58aa2d9549f5188910e2a77b438516714f.
2011-02-14graw: remove pipe_vertex_buffer::max_indexVinson Lee
This is a follow-up to commit cdca3c58aa2d9549f5188910e2a77b438516714f.
2011-02-14st/mesa: Use blend equation and function of first render target for all ↵Fabian Bieler
render targets if ARB_draw_buffers_blend is not supported If EXT_draw_buffers2 is supported but ARB_draw_buffers_blend isn't _mesa_BlendFuncSeparateEXT only sets up the blend equation and function for the first render target. This patch makes sure that update_blend doesn't try to use the data from other rendertargets in such cases. Signed-off-by: Brian Paul <brianp@vmware.com>
2011-02-14r300g: implement pb_manager::is_buffer_busyMarek Olšák
2011-02-14pb_bufmgr_cache: add is_buffer_busy hook and use it instead of non-blocking mapMarek Olšák
This is cleaner and implementing the hook is optional.
2011-02-14gallium: notify drivers about possible changes in user buffer contentsMarek Olšák
Also implement the redefine_user_buffer hook in the drivers.
2011-02-14st/mesa: set vertex arrays state only when necessaryMarek Olšák
The vertex arrays state should be set only when (_NEW_ARRAY | _NEW_PROGRAM) is dirty. This assumes user buffer content is mutable, which will be sorted out in the next commit. The following usage case should be much faster now: for (i = 0; i < 1000; i++) { glDrawElements(...); } Or even: for (i = 0; i < 1000; i++) { glSomeStateChangeOtherThanArraysOrProgram(...); glDrawElements(...); } The performance increase from this may be significant in some apps and negligible in others. It is especially noticable in the Torcs game (r300g): Before: 15.4 fps After: 20 fps Also less looping over attribs in st_draw_vbo yields slight speed-up in apps with lots of glDraw* calls.
2011-02-14gallium: remove pipe_vertex_buffer::max_indexMarek Olšák
This is redundant to pipe_draw_info::max_index and doesn't really fit in the optimizations I plan.
2011-02-14gallium: always save and restore vertex buffers using cso_cacheMarek Olšák
2011-02-14vbo: bind arrays only when necessaryMarek Olšák
We don't need to call bind_arrays in the vbo module if the states which the function depends on are not dirty.
2011-02-14vbo: notify a driver that we change buffer offsets, strides, etc.Marek Olšák
2011-02-14r300g: Remove redundant initialization.Vinson Lee
Remove redundant initialization from commit 3b01b52bd78e3d2fc857feacebd815a5fae00c94 noticed by tstellar.
2011-02-14r600g: add alignment cases for linear alignedAlex Deucher
Matches the drm and ddx. Signed-off-by: Alex Deucher <alexdeucher@gmail.com>
2011-02-14draw: Don't define the last two mipmap levels in aa lines.José Fonseca
Some hardware needs to make a seperate copy of all levels to obey max_lod, so simply do not define the last two levels instead of clamping.
2011-02-14svga: Set the appropriate flags when creating sampler/surface views.José Fonseca
2011-02-14r300g: put indices in CS if there's just a few of them and are in user memoryMarek Olšák
2011-02-14r300g: cleanup the draw functionsMarek Olšák
2011-02-14r300g: do not create a user buffer struct for misaligned ushort indices fallbackMarek Olšák
2011-02-14r300g: fix fallback for misaligned ushort indices with num vertices >= 65535Marek Olšák
2011-02-14r300g: consolidate emission of common draw regsMarek Olšák
2011-02-14r300g: Move declaration before code.Vinson Lee
Fixes SCons build.
2011-02-14r600g: do not destroy the original index buffer when translating indicesMarek Olšák
Because we only translate a subrange of the buffer.
2011-02-14r300g: emit 3D_LOAD_VBPNTR only when necessaryMarek Olšák
I thought I couldn't skip emitting this packet in some cases. Well it looks like I can.
2011-02-14r300g: fix reference counting when translating indicesMarek Olšák
2011-02-14u_index_modify: map buffers with PIPE_TRANSFER_UNSYNCHRONIZEDMarek Olšák
2011-02-14r300g: remove the relocation after AARESOLVE_PITCHMarek Olšák
2011-02-14egl_dri2: add nouveau support.Dave Airlie
but really wtf? all these PCI IDs need to be ripped out of here, its totally unscalable and the drivers already have this info so could export it some better way. tested by Darxus on #wayland.